Yesterday nite, our group decided to pull out one of the older games
we had...and chose Illuminati because of the easy rules (compared to
some of the other choices we had).
We played a few games and loved it. Couldn't decide why we stopped
playing it at all. However, in the last game we hit a snag.
The Gnomes of Zurich were absent from every game until the last. The
games before were great, with a ton of bargaining, backstabbing, and
general mayhem between the players as everyone tried to win and stop
everyone else. The Gnomes really screwed this up.
The problem is that the Gnomes can essentially sit through the game for
13 turns and be assured a win...if the game makes it that long. Also,
to shorten this 13 turn wait, they can attack low income, low resistance
groups to slowly boost their income. Most players won't block their
attempts because it would leave them short on cash to stop other players
whose possible winning conditions would be potentially shorter to
attain.
What wound up happening was that the Gnomes (a) didn't even really "play"
the game since they were doing fine by staying out of the action, and
(b) wound up "ruining" the game because other players had no way of
really stopping them, or even slowing them, on their inevitable climb to
150MB. Another player eventually won due to the fact that the Gnome player
got fed up with just sitting there and essentially "threw" the game by
beginning to attack other people's power structures that were high in
income. The player who was missed by these attacks won by collecting the
final group needed - the sixth violent for the Assassins.
Are there any rules I overlooked that can slow the Gnomes down? I read
through the book today, but didn't find anything. It just seems as if
the Gnomes can make a normally dynamic game a real bore as the Gnome player
takes a very stoic, slow burn approach in a normally fast moving, dynamic
game. We decided to ban the Gnomes in the future unless we could decide
upon an alternate winning condition.
As a final note: We're playing the OLD version of Illuminati - in all the
little plastic boxes and such...not the newer version that comes all together
in the newer box. The new version of the rules could have a different
winning condition for the Gnomes...but we don't know.
Any ideas on what to do with them?
Thanks,
Matt
mx...@po.cwru.edu
Admittedly, the main reason Gnomes don't win _as_ often [as not] in the
games I've seen is that they don't just "sit around". It is a concise
strategy. The easiest way to overcome it, unfortunately, is by consortium.
WIth, say, 4 others in the game, the opposing combined income ('nati groups
only) will be 27-37 to the Gnome's 12. And if all 'nati groups hold two
groups, a corsotium will [likely] have more income. IF everyone chips in
2MB per Gnome attack ... well, let's say I've played a fair number of games
wherein one disliked group was killed off after the third turn, since they
had *no* groups. <grin>
THe only way the Gnomes accumulate MB is if no one opposes them. And if they
stop attacking [to control], then start attacking their power structure.
Make 'em spend MB !!
-CS
To beat them you must recognize what is happening and get a coalition
going to keep the Gnomes busy defending themselves. Remember, they
have to keep at least 1 group around in order to survive. A good way
to do this is to be REALLY concerned with the status of the Gnomes.
make noises about them, sound worried, get the other players going in
that direction and you won't have any problems keeping the Gnomes in
line.
Lance
--
Lance A. Brown <br...@cs.utk.edu>
The Crystal Wind is the Storm, and the Storm is Data,
and the Data is Life -- The Player's Litany
from _The Long Run_ by D.K. Moran
I agree with this. Since the original poster said that this was his group's
first game with the Gnomes, I would guess that the players just didn't
realize that they need to treat the Gnomes differently than other Illuminati.
Don't be so quick to change the game rules, change your strategy first
and see if there is still a problem.
--
Ken Yousten "That's one of the tragedies of life- the men
Blacksburg, VA most in need of a beating-up are always enormous"
you...@csos.orst.edu Rudy Vallee, "The Palm Beach Story"
This player, let's call him Bruce (hi Bruce :^), kept detailed notes of
exactly how much money the Gnomes had at any given time. Every time they
earned income or spent any money, Bruce jotted it down on a piece of paper,
so everyone knew exactly how well (or poorly) they were doing... And
whenever Zurich got too high in the minds of the other players, they'd just
take turns whacking off groups until the Gnomes were forced spend money to
defend themselves (or be eliminated, by the at-least-one-group rule).
This was very effective at stopping the Gnomes, but very frustrating for the
Gnomes' player, so much so no one in that group ever chose the Gnomes again.
Unfortunately, this sort of extreme tactic seems necessary against the
Gnomes--and against another unpopular (in our group) Illuminatus, the
Servants of Cthulhu.
The problem with Cthulhu is that there's no way to pull them further from
their goal. Any of the other groups, can be dragged back from their victory
conditions, by forcibly depriving them of whatever it is they're trying to
accumulate. There's no way to un-kill a group, though, which means that to
stop Cthulhu, you either have to win before they do, or eliminate them from
the game. Because the other three to five players are competing to do the
former, but can cooperate to do the latter, we've often found Cthulhu
getting eliminated as soon as they kill their 6th or 7th group. (They need
eight, right? It's been a while; adjust numbers accordingly if necessary.)
Needless to say, that tends to make Cthulhu a lot less fun to play, though
clever players in our group overame this problem by simply not destroying
any groups, instead going for the base-condition win by trying to build a
strong power structure. It's not as easy as the other way, but it's a lot
more satisfying than knowing that as soon as you get close to winning, the
other players are going to have no choice but to kill you.
I still have mixed feelings about Illuminati--I have a hard time playing a
game that's both intense and serious in terms of alliances and backstabbing,
but at the same time very fast, random, and general silly. If I want
intense, I'll play something less random, like Diplomacy; if I want beer-
and-pretzels, I'd rather go with something like Wiz War or Cosmic Encounter.
Your mileage, however, may vary.
--
Josh Smith, User Support Coordinator :: Pleasure: iri...@hmc.edu
Harvey Mudd College, Claremont CA USA :: Business: jo...@hmc.edu
From the hedge-row's fragile cover, I saw brother killing brother,
And all of this was done in Jesus' name.
>Yesterday nite, our group decided to pull out one of the older games
>we had...and chose Illuminati because of the easy rules (compared to
>some of the other choices we had).
>We played a few games and loved it. Couldn't decide why we stopped
>playing it at all. However, in the last game we hit a snag.
>The Gnomes of Zurich were absent from every game until the last. The
>games before were great, with a ton of bargaining, backstabbing, and
>general mayhem between the players as everyone tried to win and stop
>everyone else. The Gnomes really screwed this up.
>The problem is that the Gnomes can essentially sit through the game for
>13 turns and be assured a win...if the game makes it that long. Also,
>to shorten this 13 turn wait, they can attack low income, low resistance
>groups to slowly boost their income. Most players won't block their
>attempts because it would leave them short on cash to stop other players
>whose possible winning conditions would be potentially shorter to
>attain.
....
1) It's not 13 turns, it's 9. Anyone may pass their turn and
collect 5 MB. Assuming that the Gnomes have a zero income side group,
17MB/turn takes 'em 9 turns to win. On the other hand, give the
Gnomes the Mafia and Video Games and we are talking about five turns
to win. *That's* tough to beat. Of course, give anyone the Video
Games and the Mafia and they're going to be tough to beat.
2) 9 turns is a long time. It's pretty unlikely that the game will
still be going on by then; anyone else can win in about five turns
if unopposed. Even five turns is too long to stay isolationist
and hope the game is still in progress when you emerge.
3) There are dozens of ways to counter that Gnome strategy. Kill
their group. Or just neutralize it. If that happens, they die.
Threaten to let someone else win. "He's going for the win. Guess
what, Gnomes, you are going to have to spend to stop him or lose."
Get the Assasins to perform community service by draining his cash
reserves by attacking to neutralize his groups. (The Assassins get
+10 gross to neutralize.) Everyone chip in 5 MB on the attack and
the Gnomes have to spend 20 or so to counter it. Do this every
other turn or so and the Gnomes had better get moving or end up
in last place in a hurry.
4) Get more interesting players. The Gnomes are great fun run with
an active strategy.
5) That which prevents players from achieving the general victory
conditions is that there are a limited number of groups available
per turn (a little less than one per player per turn unless the
Network is in play.) If one player is not attempting to control
any groups, the others will quickly move sufficiently ahead that
they will each be threatening to win.
6) Even the UFOs with the Mafia will have a larger income than the
Gnomes. If the Gnomes just sit back, others will outproduce them
by large factors within a few turns, unless only garbage comes out
in the center. If the rest of the players are outproducing the
Gnomes, they will have no problem killing all their groups and
KOing them from the game.
7) Which is why the grand isolationist strategy for the Gnomes is
not sensible. If it doesn't work the first time, they lose the
game for sure because the counterstrategy kills them dead.
--Jeff
PS: On the other hand, the Bavarian Illuminati are almost unbeatable
unless the other players gang up to delete them from the game around
turn three. Imagine always having a card in your hand. They make
the Network look like children.
--
Here's a half a dollar if you dare
Double twist if you hit the air.
Look at Julie down below
The levee doin' the do-pas-o.
We always found the Bavarian Illuminati to be the toughest ones to beat.
--
- Bill Seurer Language and Compiler Development IBM Rochester, MN
Business: BillS...@vnet.ibm.com Home: BillS...@aol.com
>The group I played Illuminati with at Swarthmore actually had the opposite
>problem with the Gnomes: no one ever wanted to play them, because everyone
>would always gang up on them. The strategies others have described make it
>virtually impossible for the Gnomes to succeed, especially when coupled with
>a particularly nasty tactic someone came up with before the last game in
>which someone voluntarily played the Gnomes.
>This player, let's call him Bruce (hi Bruce :^), kept detailed notes of
>exactly how much money the Gnomes had at any given time. Every time they
>earned income or spent any money, Bruce jotted it down on a piece of paper,
We have a "house rule" that this tactic is illegal. We use that
rule for all games in which something can be kept track of but is
expressly given in the rules as not being "public knowledge." For
games in which it is trivial with a modicum of effort (and no paper)
to keep track of such things yet the rules claim secrecy, we remove
the secrecy and make it countable at need.
>The problem with Cthulhu is that there's no way to pull them further from
>their goal. Any of the other groups, can be dragged back from their victory
>conditions, by forcibly depriving them of whatever it is they're trying to
>accumulate. There's no way to un-kill a group, though, which means that to
>stop Cthulhu, you either have to win before they do, or eliminate them from
>the game. Because the other three to five players are competing to do the
>former, but can cooperate to do the latter, we've often found Cthulhu
>getting eliminated as soon as they kill their 6th or 7th group. (They need
>eight, right? It's been a while; adjust numbers accordingly if necessary.)
>Needless to say, that tends to make Cthulhu a lot less fun to play, though
>clever players in our group overame this problem by simply not destroying
>any groups, instead going for the base-condition win by trying to build a
>strong power structure. It's not as easy as the other way, but it's a lot
>more satisfying than knowing that as soon as you get close to winning, the
>other players are going to have no choice but to kill you.
I've never seen the Servants win with their special victory condition.
It's really not sensible to try too hard to get there. If you do try,
you must get to six and stop. Then get two cards and try to win in one
turn. Ensure that no one knows that you have two cards, though that is
not very easy to do unless the game is very very long. Alternatively,
use the "Servents' Sleaze:" Get to five deaders, kill one group as your
first action, release all others but one as your second, then kill your
own last group, killing your illuminati as well. That makes eight and
a win. It takes some players by surprise. Also, if the game is long,
which ours tend to be, the number of groups killed by the Servents may
get forgotten if they lay low for a long time. Overall, though, the
Servants are not a winning Illuminati.
I rate the Illuminati roughly:
Bavarians
Network
UFOs
Gnomes
Bermuda Triangle (Higher without the OMCLs)
Assassins
Servants
Discordians
The Bavarians win almost always if they are in the game (or lose
early.) The Network is very powerful and usually wins the rest of
the time. The UFOs, Gnomes, and Bermuda have a decent chance to win.
The Assassins have an outside chance, and the Servants and Discordians
are very very weak.
But all the players in our group know this, so preventive measures
are taken. If the Mafia shows up, we might all contribute to the
Network's not getting it, then proceed not to spend a penny against the
Discordians. We try to balance the group's power with this sort of
favoritism, so after awhile, the powers sort of balance out. The
Network's ability to get gobs of cards is hard to beat, though, and
the Bavarian's 5MB for a priviledged attack is impossible to beat, so
they still win more than their share. In our games, however, not the
Bavarians. We don't use 'em. No one likes to be eliminated turn three,
and that's the only way to handle the Bavarians.
--Jeff
I want to try to play a game of Magic by EMail. I'm trying to
compile some ground rules. I'll throw out some ideas for the
net to consider. Discussion is welcome. I will try to compile
the comments I receive here on the net and via Email and post
a summary.
I would also be looking for one or two people to try these rules
out against. But I would probably only be able to handle two games
at once.
1. The honor system will apply. Obviously, since I will have my
deck here and you will have your deck there. Players will be
asked not to "stack" their deck or use cards they do not own.
2. How to play the decks?
a. If a player had a deck of cards that he wouldn't mind
setting aside for a few months, he could simply shuffle
those cards and set them on a shelf somewhere, drawing
cards as needed.
b. If, like me, a player wants to use many of the cards for
real life play also, some way of "substituting" cards is
needed. I would suggest using a deck of regular playing
cards (or two different-colored decks if you put more than
52 cards in your play deck). On a piece of paper, or on
a computer file, you list your cards opposite the playing
card that represents each card (i.e., the Jack of Diamonds
is a Llanowar Elf). Then, shuffle the playing cards and
when you draw a card, you look it up on your list and play
accordingly. Those of us with programming ability could
probably write a program to shuffle the cards.
c. Duplicate cards should be given a number as they are
played, such as, "I play Hill Giant #2".
3. What deck-building rules to use? Personaly, I like the
tournament rules that seem to have been leaked from WoTC. In
a nutshell, they are; 60 cards minimum, 5 Artifacts maximum
(no duplicates), and 4-of-a-kind maximum for other spells. But,
of course, any limits any two people set up is fine.
4. Ante? If playing for ante, the players would have to agree to
mail the card to the winner. Again, the honor system would apply.
5. Physical play. I think the Chaos Orb is the only card that
requires the physical presence of the cards. I would suggest
that the play of this card requires a pair of 6-sided dice to be
rolled for each card the opponent has in play. On a roll of 11
or 12, the card is sent to the graveyard. Other cards (like
Natural Selection) would have to be dealt with when they happen.
6. Frequency of Email. I would be looking for a person that could
perform two moves a day by Email. Other people could go for one
move a day.
7. Order of Play. I would guess that simple language should be used.
A sample turn might be,
'I draw a card. I play "Forest #4". I tap Islands #1,2,3 to
Summon Prodigal Sorcer #1. I use Icy Manipulator #1 to tap
your Wall of Ice #1 and attack with my Senguir Vampires #1,2.'
Your opponent could then use Instants and Interupts to void
parts of your turn. You would then have an opportunity to
replay the rest of your turn from the point of Interuption.
'After you play Forest #4, I tap Mountains #5,6,7 and Shatter
your Icy Manipulator.'
This method has the disadvantage of the attacker revealing his
plans before he knows whether they can be completed. But, I
would guess that, for especially sneaky moves, you could submit
the parts of your turn in smaller pieces. Otherwise, I would
say that that's one of the sacrifices you have to make to make
PBEM Magic playable.
>
> I've never seen the Servants win with their special victory condition.
> It's really not sensible to try too hard to get there. If you do try,
> you must get to six and stop. Then get two cards and try to win in one
> turn. Ensure that no one knows that you have two cards, though that is
> not very easy to do unless the game is very very long. Alternatively,
> use the "Servents' Sleaze:" Get to five deaders, kill one group as your
> first action, release all others but one as your second, then kill your
> own last group, killing your illuminati as well. That makes eight and
> a win. It takes some players by surprise.
Actually, this is expressly forbidden in the new rules, but it made for
a tremendous ending the last time anyone won with the Servants.
Scenario:
Servants killed after 6 destroyed -- with a lot of paranoia.
Servant player hangs around carping. Eventually, one player reveals
he holds the Media Campaign card and threatens to bring Cthulhu
back, playing on extreme Servant paranoia of other players.
Servant player, through impressive rhetoric, convinces player to bring
him back, "To distract his opponents."
Servant takes Democrats and uses them (with resurrecting player's help!)
to get kill 7. Two actions complete.
Servant freely drops Democrats (as can be done at any time), killing self
for win.
There may be rule violations you can drive a truck through, but the
Servant took _everyone_ by surprise, and was voted the win anyway for sheer
chutzpah.
>
>
> The Bavarians win almost always if they are in the game (or lose
> early.) The Network is very powerful and usually wins the rest of
> the time. The UFOs, Gnomes, and Bermuda have a decent chance to win.
> The Assassins have an outside chance, and the Servants and Discordians
> are very very weak.
>
Actually, our group hates the Bavarians so much, they have to work extra
hard -- no one likes to let them even get _near_ striking distance, and are often
ganged up on. The Network generates a similar reaction (plus we upped the
deferred requirement to 18. 15 was way too easy). UFOs and Triangle (followed
by Weirdians -- usually overlooked) are our most-win groups.
Glad to see some I! discussion . . . JJMcC
: We (by consensus) removed the Gnomes from the mix for just that reason.
: Hated to be 'em, hated to play against 'em because it invariably neutralized the
: player preceding the Gnome as well. It seems all the other players refuse to
: use actions to deal with the gnome, requiring the last slob to be in perpetual
: 'Gnome Attack' mode and never progress on his own.
The thing for this poor slob to do is to threaten to do absolutely
nothing! This will bring concessions (like donations of MBs, power
transfers, cards, etc) from other groups very quickly. And the best
part? He cannot be stabbed, because he is needed to to attack the Gnome
the next turn.
I think Ill. is one of my favourite games for the past 10 years because
of its intense politiking going on at all times. Sure, the Bavarian 5MB
privilaged attack is hard to beat, but threats of retailatory attacks by
a consortium will usually keep such actions to a minimum.
It bears keeping in mind that although some groups are stronger than
others, few are as strong as two and none can stand up to three others.
So take heart! Even that pesky UFO can win by keeping others guessing of
its own victory conditions, and volunteering to do community services
(some of which has absolutely no relevance to your victory conditions)
Because others are wary of each other, they usually forget you until too
late...
Sometimes, I think in terms of intrigue, it is better than even Diplomacy!
Just my thoughts.
-Titan
****************************************************************
* Standard Disclaimers * Please post any comments and flames *
* Apply * to eng2...@leonis.nus.sg *
****************************************************************
* "I find your lack of faith disturbing..." *
* - Darth Vader *
* Star Wars *
****************************************************************
I have noticed an option that is not much discussed in this thread,
and that is joint victories. When I was playing Illuminatti regulary
I think there were only 2 games I played that had an individual
winner. In one the Network drew something like 6 special cards in her
first 4 turns:-( In the other we were using the optional rules for
illuminated groups. The Post Office has a special power to force a
player to skip his turn. 2(!:-) players still had no groups at the
beginning of the third turn. The PO took bribes to use his special
power on the other, then excepted BOTH bribes and basically turned the
game into a 3 on 1 with the PO guarenteed to win. In every game we
played two people would arrange a swap or trade at the end of a turn
to give both a shared victory. In this case the Gnomes are
everybody's friend since everybody has something they want, cash.
When you look the special victory conditions, and special powers you
need to divide the Illuminatti into 3 groups. Those whose powers and
conditions give positive feedback to each other. Bavaria is the best
example of this; the closer to the victory condition the more power
they have so it is easier to get closer. The second group are those
where gaining your special victory condition at least strengthens your
position. The collector groups are in this category. By collecting
groups they imporve their position, but there is no feedback. The
Gnomes can fall in either category. They get extra income and win
with the total income, but to use the income they move AWAY from their
victory condition. The last group are those with negative feedback.
Cthuthlu is the best (only?) example. In this case, if they try to
work towards their special victory condition, they are in essence
giving up a turn to strengthen their position. Even worse, destroying
a group adds nothing to their position and might even consume
resources.
Most of the people I play with are competitive enough that they regard a shared
victory as not really a victory at all, so this never happens. This sort of
rule is my big gripe about the game Dune, where shared victories are about the
only way to win. Illuminati encourages enough distrust to keep shared win
sliminess to a minimum, however.
My $0.02 on the Gnomes debate - I've only ever seen the gnomes win via the $150
rule once - I was the gnomes, and people ignore me because I had such a small
pile of money (1 of which was actually a $100 bill :). The reason it's hard
for the gnomes to win this way is that to flex their power, they must spend the
money, which means spending their victory points. By the time that the gnomes
have a power structure large enough to avoid getting waxed by a couple well
placed attacks to neutralize (and thus are in a position to start saving their
money), there are inevitably one or more other powers that are close to winning
and thus require defensive action that eats into the gnomes treasury. All this
is apart from the fact that a gnome who holes up and saves gets whaled on.
I agree with the Bavarians being too powerful. Has anyone tried raising their
win requirement by 5 power or so to see how that helps? Otherwise they simply
must be exterminated, which is no fun for the player.
Group that most often wins : UFO's - you never know exactly what they're trying
to do, and there's always someone else that you can see and be sure that needs
clobbering.
My favorite group - Assasins. Their power lies in being the favorite implement
of all the players to knock the leader down a notch. Fun Fun Fun. They are
also the best at my favorite tactic to deal with players who have too much
money - the random attack to neutralize with no hope of actually succeeding.
Just use that spare action to make the target spend a few $ to be safe.
Alec
>In article <2fp9sh$3...@newsflash.mitre.org> ter...@smiley.mitre.org (Terry L Rooker) writes:
>>In article <1993Dec27.1...@rchland.ibm.com>,
>>I have noticed an option that is not much discussed in this thread,
>>and that is joint victories. When I was playing Illuminatti regulary
>>I think there were only 2 games I played that had an individual
>>winner.
>Most of the people I play with are competitive enough that they regard a shared
>victory as not really a victory at all, so this never happens. This sort of
>rule is my big gripe about the game Dune, where shared victories are about the
>only way to win. Illuminati encourages enough distrust to keep shared win
>sliminess to a minimum, however.
We, too. We have institutionalized this feeling and prohibit shared
wins. In the case that it occurs accidentally, we call it a tie and
everybody loses. (Heh, heh.)
>I agree with the Bavarians being too powerful. Has anyone tried raising their
>win requirement by 5 power or so to see how that helps? Otherwise they simply
>must be exterminated, which is no fun for the player.
I don't know about raising their special victory condition, but I like
the idea of modifying them ok. How about making their priviledged attack
cost 10 MB? Or 15? That'd bring them down to size pretty fast. In fact,
it's such a good idea that I'll suggest it next time we play.
>Group that most often wins : UFO's - you never know exactly what they're trying
>to do, and there's always someone else that you can see and be sure that needs
>clobbering.
For us, it's the network. Those cards are really good. The UFOs are
lots of fun, too. They are always doing something, either trying to
help themselves or trying to create red herrings. Destroy those groups,
snag those weirds, what are they up to!?
>My favorite group - Assasins. Their power lies in being the favorite implement
>of all the players to knock the leader down a notch. Fun Fun Fun. They are
>also the best at my favorite tactic to deal with players who have too much
>money - the random attack to neutralize with no hope of actually succeeding.
>Just use that spare action to make the target spend a few $ to be safe.
Also fun, but there's almost always something to do with an action.
When I'm the Assasins, I try to get other players to fun my soak-off
attacks so that it'll really cost the leader some bucks. In a six-player
game, that's an additional eight bucks for only two MBs spent each. I, of
course, "contribute" my action. (Heh, heh. Wring those hands...)
They have no hope of winning if they are trying to collect the same
underlings as another group. It is hard enough to collect N
weird/violent/power transfer groups without having two groups do it as
their primary goal.
I disagree. Early on, if you headfake in a number of different ways,
you can accumulate quite a few at the expense of the 'real' collector -- "Just
so the Weirdians/Assassins/Triangle/etc doesn't get 'em." The UFOs in our games
win just as much when they match an opponents' goal as not. Sounds like you just
need more subtle UFOs ;]
----------------------------------------------------------------------------
|
Jeff J McCoskey | "Preservatives might be preservin' you all:
DoD# 750A2 | I think that's somethin' you mighta missed."
| -- Jefferson Airplane
----------------------------------------------------------------------------
Editor-In-Chief: Pseudo-Random House
publishers of the _Agent of PULP_ series on alt.comics.lnh
----------------------------------------------------------------------------
je...@gg.caltech.edu (Jeff Goldsmith) writes:
>We have a "house rule" that this tactic is illegal.
Hurm... I know people who could and would do this in their head
just as insurance. The Gnomes were the first and last group I ever
played, though there was about 4 years in between the two games. *grin* I
think the guy we're calling Bruce (hi Bruce!) laid off me in my first game
'cos I was a newbie, though it didn't really matter. (lost horribly...)
Someone (Not this mythical Bruce creature....) did count my MB in my last
game, however, and completely ruined the game for me. I tried to be tricky,
even resorting to giving money away, to no avail. Kinda put me off the
Gnomes forever... Perhaps if there were some way for the Gnomes to slightly
randomize their money? (Short of putting a random chit back into the
bank... Hum... That could be interesting...)
je...@gg.caltech.edu (Jeff Goldsmith) writes:
>I rate the Illuminati roughly:
>[...]
Just a random observation from watching the net over the years.
Odds are some other group has a different ranking (witness Titan, A&A, etc.)
than yours only because of a different playing style. Wouldn't a traveling
circut of gamers be cool? Or we can just wait for the Information Highway
(tm) to make all these games simple to play over the net.... Yea....
-Luke
--
______________________________________________________________________________
Luke Hankins han...@cs.swarthmore.edu |"Le coeur a ses raisons que la raison
for(MyAge=0;!Dead;EnjoyLife); | ne connait point" -Blaise Pascal
While not (EndOfLife) do Havefun(me); |"...stay sane inside insanity..." -RHPS
I'd like to play but wasn't sure if I found the right address to respond by
email. I'm dur...@cs.uoregon.edu
> 1. The honor system will apply. Obviously, since I will have my
> deck here and you will have your deck there. Players will be
> asked not to "stack" their deck or use cards they do not own.
I'd like to see a provision for those of us that have the Collector's
Edition and nothing else (yet). Set up rules for composing decks assuming
that any card is available. Something like the tournament rules you
mentioned. Given the set-up in 2b (using other cards to represent MtG
cards with a database of the mapping) that seems relatively easy. Of
course, it eliminates the trading-card aspect of the game, but since that
relies on the honor system anyhow, it might be a reasonable approach for
PBEM.
> b. If, like me, a player wants to use many of the cards for
> real life play also, some way of "substituting" cards is
> needed. I would suggest using a deck of regular playing
> cards (or two different-colored decks if you put more than
> 52 cards in your play deck). On a piece of paper, or on
> a computer file, you list your cards opposite the playing
> card that represents each card (i.e., the Jack of Diamonds
> is a Llanowar Elf).
This is also a good way of incorporating Collector's Edition and non-CE
cards in the same way. It might be worth setting up a small program to
facilitate it.
Brian Durney dur...@cs.uoregon.edu
who thinks that WotC shouldn't have bothered with the Collector's Edition
and if they did they should have said "WARNING this box does not come with
a set of the rules and isn't meant to play with, just to look at, even
though these are *supposed* to be *playable* trading cards" in big, red
letters across the top. :-(
>Group that most often wins : UFO's - you never know exactly what they're trying
>to do, and there's always someone else that you can see and be sure that needs
>clobbering.
In my groups the other player get my real intentions really fast. If you
choose some exotic thing they can 'see' it very fast. (Normally after trying
to get some 'shit' card that noone wants :-))
Heiko
PS: I think Illuminati is with Grass the best beer and pretzel game.
Well, if you are ever in Pittsburgh, I'm always looking for a few players...
-Brian
-------------------------------------------------------------------------
| "Are they dead?" |
| "Does it matter?" |
| - Pugsley and Wednesday in "The Addams Family." |
-------------------------------------------------------------------------
Admittedly most of our games are fairly paranoid. A solo win by the
Discordians when everyone had just expended all their effort stopping another
player from winning, and a solo by the Bavarian Illumnati using their
special ability are the only lone wins to date. The rest have been 2-way wins
at best,
John
Most of the people I play with are competitive enough that they regard a
shared victory as not really a victory at all, so this never happens.
This sort of rule is my big gripe about the game Dune, where shared
victories are about the only way to win.
I strongly disagree about your assesment of Dune. I have played this
game *many* times and have seen more solo victories than alliance
victories, although admittedly alliance victories are still around
1/4 to 1/3 of the outcomes. One of my favorite solo vitories was actually
through an alliance: I convinced everyone playing to ally against one other
player who had been really annoying, so that he would be the only one to
lose. However, I was playing the BGs and had predicted one of my allies
to win in that turn, so I got the solo victory instead.
- Chris
--
Christopher Barber
(cba...@bbn.com)
Ming
--
Eu-Ming Lee mi...@interaccess.com
"i showed this girl my stitches
"she said she had some too
"she said she thinks she'll start a rock band too." - Throwing Muses
>iri...@hmc.edu (Josh Smith) writes:
>>This player, let's call him Bruce (hi Bruce :^), kept detailed notes of
>>exactly how much money the Gnomes had at any given time. Every time they
>>earned income or spent any money, Bruce jotted it down on a piece of paper,
>je...@gg.caltech.edu (Jeff Goldsmith) writes:
>>We have a "house rule" that this tactic is illegal.
> Hurm... I know people who could and would do this in their head
>just as insurance. The Gnomes were the first and last group I ever
>played, though there was about 4 years in between the two games. *grin* I
>think the guy we're calling Bruce (hi Bruce!) laid off me in my first game
>'cos I was a newbie, though it didn't really matter. (lost horribly...)
>Someone (Not this mythical Bruce creature....) did count my MB in my last
>game, however, and completely ruined the game for me. I tried to be tricky,
>even resorting to giving money away, to no avail. Kinda put me off the
>Gnomes forever... Perhaps if there were some way for the Gnomes to slightly
>randomize their money? (Short of putting a random chit back into the
>bank... Hum... That could be interesting...)
Perhaps, but there's a simpler way. "Bruce" is not permitted to
share his information. He's welcome to count, but he cannot use
paper or other aids, and he's not allowed to tell anyone. "I think
the Gnomes are getting close" is as far as he's allowed to go. If
he says, "The Gnomes have 143 MB," then he either gets ostracized from
the group, or he is made to pay a penalty (20 MB, say) for "slander,"
since this tactic makes many games far less playable. An alternative
is to tell Bruce, "That comment requires you to do three shots of something
vile right now." Let him count after a couple of penalties like that! :)
The idea of making the Gnomes' income random is a good one, though.
It makes it difficult for them to plan exactly to the dollar how much
money they will have. The problem is keeping this information secret.
One can either trust the Gnomes (yeah, right) or have them show one
person the dice roll. If it's someone different each turn, no one will
have complete information. How about changing their income to 7+2D4?
That's a range of 9-15, which, while not a big difference, should be
enough to foil counting.
>In my groups the other player get my real intentions really fast. If you
>choose some exotic thing they can 'see' it very fast. (Normally after trying
>to get some 'shit' card that noone wants :-))
Don't go after those cards, then. Or go after two of them, one
being useless to you. Usually, I find that it's best to pick
a victory condition with the UFOs that some other player is
attempting. You win by making a priviledged attack on that
players power structure, taking the groups that they've carefully
collected and you really need.
The real bonus of the UFOs, though, is that they get to use their
transferable power twice per turn. That's like an additional six
income. If they get two high-power groups, they are going to be
tough in combat. "Let's see, The Mafia attacks with power from
Madison Ave, and, of course, the six from the UFOs. Like Violence
makes that a 21 base. Anyone want to try to stop them? Ok, for my
second attack..."
>The idea of making the Gnomes' income random is a good one, though.
>It makes it difficult for them to plan exactly to the dollar how much
>money they will have. The problem is keeping this information secret.
>One can either trust the Gnomes (yeah, right) or have them show one
>person the dice roll. If it's someone different each turn, no one will
>have complete information. How about changing their income to 7+2D4?
>That's a range of 9-15, which, while not a big difference, should be
>enough to foil counting.
What you could do is take a deck of ordinary playing cards. Each turn, the
Gnomes may take $12 or draw 2 cards from the shuffled deck. Count aces as '1'
and face cards as '10'. These cards count towards the gnomes victory condition.
The gnomes can cash in any of the cards at any time if he needs the money for
something. Also allow collecting 1 card and $6. This will give the Gnomes an
extra $.54 per card drawn on average. If deck runs out, reshuffle. If no cards
left, gnomes should have won as the deck represents $340. Money from cards
are assumed to be on the illuminati itself.
--
===============================================================================
Geoffrey DeWan -- gde...@mrcnext.cso.uiuc.edu
I must admit, I've played at least 100 games of Dune over
15 years, and in that time I've only seen maybe <10 solo
victories, all by sheer fluke. (Except for the 4-player-no-alliance
games we've played.) Of course, to put that in context, nearly all of
those games were with the same group of guys, and I wouldn't
swear that our style of play is normal...
First of all, we _always_ play with a house rule limiting
formal alliances to 2 players only. The normal flow of play
is: Jockey for position with minimal conflict for 2-3 turns
(on average) before the first worm appears. Then the 2 strongest
factions (almost always involving the Harkonnen) form an alliance
intending to go for the quick victory, and the other 4 factions
form into 2 formal alliances, but informally acting as an alliance
of 4 against the leaders. Then a 1 turn BLOODBATH ensues. When the
dust clears, at least 2 factions are almost completely in the
tanks (almost always involving the Guild [Policemen of Dune])
and one of the two strongest 2-faction alliances have won. If
no winner emerges, that usually means that 3 or 4 factions are
almost entirely in the tanks, and whoever has the most tokens
left will usually sleaze into a win the next turn. Despite the
relative predictability of this type of game, BG predictions are
quite rare, primarily due to the fact that you can't tell when
the first worm will appear. (The favorite prediction is
"Harkonnen in 4".)
We've tried to play with the "official" alliance rules (unlimited
number of partners), but this just became absurd -- How can the
strongest 3 (or more) factions _not_ control 3 cities at the end
of the turn? They'd have to be either really unlucky or really
foolish, IMHO. As you pointed out in your note, the increased
chance of the BG prediction being correct should deter this
type of play, but in the few times we tried it, it didn't.
When our alliances occur, since they normally mean the end of
the game, the 2 allies share EVERYTHING (all traitors, cards, spice,
etc) without question. The Atreides ally looks over his shoulder
to see the cards up for bid. The Emperor ally gets his treachery
card hand filled for free. No questions asked. The more information
available, the more likely the alliance can win. Add this to the fact
that we have no stigma in our group against joint wins. Given all of
that there is absolutely no incentive in the mechanics of the game
to attempt a solo win.
To extend the game and make solo victories more possible (and
attractive), we've tried house rules such as:
1) _Forbid_ alliances before turn X (usually X = 3-6).
2) _Require_ all alliances to be dissolved (forming
all new ones is allowed) at _every_ worm.
3) Require alliances to control 2+X cities to win
(where X = number of players in alliance).
Using one or more of these rules, we have had a game last
as long as 12 turns. That was previously unheard of.
(A 6-turn game is normally absurdly long for us. However,
playing 4-player-no-alliances, the normal ending amazingly
usually goes to either the Guild or Fremen default win.)
I could discuss Dune all day. Don't misunderstand my comments
above -- Dune is one of my personal top 5 games of all time
(along with Diplomacy, Civilization, Republic of Rome, and
History of the World). If would like any elaboration about
our style of play, just ask. Can you fill me in on the
details of your strategies and styles?
Brad Johnson (joh...@cig.mot.com)
Brad Johnson <joh...@ferret.cig.mot.com> :
> I must admit, I've played at least 100 games of Dune over
> 15 years, and in that time I've only seen maybe <10 solo
> victories, all by sheer fluke.
I think it depends a lot on how the players assess a shared victory.
If everyone just tries to maximize their chances to be one of the
winners, shared victories will of course be much more common than if
they just try to maximize their chance to be the solo winner.
Unfortunately (?) it's often unspecified in games what you really
should aim for, except for obvious things like "winning is better than
losing".
In games with a possibility of shared victories *I* often assess a
solo victory as infinitely better than a shared victory, so only when
my chances of a solo victory is nil I will try to win a two-part
victory, which in its turn is infinitely better than a three-part win,
etc.
Another view which I find very reasonable is to view an n-part win as
worth 1/n of a solo win. It's as if there was a reward for the
winners of the game to share equally and you're trying to maximize
the expected value of your share of the reward.
What do you think about positions in games where you finish first,
second, third, etc? I do it in a similar way to the shared victories,
and assess a gold medal as infinitely better than a silver medal etc.,
and will thus run a risk of finishing last for a slight chance at
winning, rather than being sure of coming in a close second.
-- "
Per Starback, Uppsala, Sweden. email: star...@minsk.docs.uu.se
"Life is but a gamble! Let flipism chart your ramble!"
: Most of the people I play with are competitive enough that they regard a
: shared victory as not really a victory at all, so this never happens.
: This sort of rule is my big gripe about the game Dune, where shared
: victories are about the only way to win.
Dune! Squeeeeeeeze them! Sorry.
In the many games that I have played (with 3 different groups of people),
3-player (or more) alliances never happened. I'm not sure why, but I think
everyone just thought it would make the game too easy to win (or lose).
2-player alliances are very workable, though. One good house rule is to
increase the number of strongholds needed to win as part of an aliance.
Since 2-player alliances were the norm, the number 4 worked quite well.
In most of the games I played, when one player or one alliance was close
to victory, the other players (or alliances) would work against the
potential victor. With experienced players, it is hard to pull off a
surprise win:
"The Hajr card hasn't been played yet. If the Guild has it, we're toast.
Look where he's got troops. We gotta bump off his ally Harkonnen.
You better take Carthag since you have the best chances of holding it
against Harkonnen. I'll take Tuek's Sietch and hope the Emperor takes
Arrakeen..."
You get the idea. Most games I have played have gone to turn 12-15, even
with alliances. So, while solo victories are rare, the game is still
quite interesting.
Mark
Actually, for alliances in Dune, we play a quirky house rule... You can form
alliances freely, but alliance wins are disallowed, i.e. YOU have to have 3
strongholds to win. It makes deciding to join an alliance a much trickier
position since you can't attack your allies' holdings. And usually your "ally"
isn't too keen on giving you the win...
Try it sometime. Add "The Duel" to it and see how many times you get an all
out leader war in the arena because no alliances are formed. ;-)
Cheers,
Roger
--
Roger Leroux | "Raise your can of beer on high,
rler...@ra.uvic.ca | And seal your fate forever.
----------------------+ Our best years have passed us by,
Be seeing you. DoD#43 | The Golden Age of Leather." -- Blue Oyster Cult
"ok, do you have a shield"
followed by
"you will play a shield defense"
when the alliance has a poison attack card, for instance. Fill in your own
permutations of the above, it makes it easy for the alliance to win if they are
moderately clever with their questions (esp. if the Atreides have kept track of
which cards you've bought).
Alec
>The idea of making the Gnomes' income random is a good one, though.
>It makes it difficult for them to plan exactly to the dollar how much
>money they will have. The problem is keeping this information secret.
>One can either trust the Gnomes (yeah, right)
My philosophy is that if you don't trust the people you play with then
don't play with them.
Of course, it's always better not to tempt people (not to mention that it's
also nice to be able to check for honest mistakes), but if it's the only
simple way to make a game go smoothly, then trust them.
>or have them show one
>person the dice roll. If it's someone different each turn, no one will
>have complete information. How about changing their income to 7+2D4?
>That's a range of 9-15, which, while not a big difference, should be
>enough to foil counting.
How about getting a set of extra money chits and put 100 chits totalling
300 money in a tin. Then allow the Gnome player to draw 4 chits every
turn. Depending on what mix of chits you choose you'd get any desired
range of money with an average of 12. For example, 50 5-chits and 50
1-chits would give you 4-20 money per turn. 20 5-chits, 40 3-chits, and
40 2-chits would give you 8-20 money per turn, but skewed towards the
low end. Or make it 40 3-chits and 60 2-chits and let the Gnomes draw
5 chits for a range of 8-15. Etc. etc. You can tailor this idea to fit
any desired range and spread.
Hans Rancke
University of Copenhagen
ran...@diku.dk
------------
"Kettelman bristled. Nothing got him angrier than when
people implied that he was paranoid. It made him feel
persecuted."
--Robert Sheckley
I've seen very few games in which it is explicitly stated that there is
such a thing as 1st, 2nd, 3rd, and so forth. Usually, the rules will say
that barring a sudden death victory, the game is won by the player with
the most victory points. In any case, people usually play this way, which
I believe is in line with how Mr. Starback sees things.
Personally, this is one aspect of multi-player games which I would just
as soon change. For one thing, it tends to turn a finely-balanced multi-sided
struggle into a mad scramble toward the end, making alliances increasingly
infeasible as the game goes on. For historical games, this is a disaster,
as it turns an ongoing nonzero-sum situation into an unrealistic zero-sum
"end of the world" scenario. Even for nonhistorical games which attempt
to simulate diplomacy, it has a distorting effect that I find unpleasant.
Another, related, problem, is that it exacerbates a typical multi-player
game phenomenon: the bash-the-leader syndrome. Toward the end of the game,
the only feasible alliances become negative ones designed to hurt whoever
is in first place at the moment. Further, the situation can often be
manipulated so as to put one player in the position of being the only
person capable of knocking off the top dog, an incredible burden which
usually results in that person either ruining their position to save the
game for everyone else or simply throwing the game to the leader.
I realize this can be a subtle tactic. (It is a highly recommended
approach in Republic of Rome.) In general it is best used against the
person who has most to lose if the leader goes unmolested, i.e., the
second-place person. However, I think it is inappropriate in many
situations. For example, in a game of international diplomacy like Pax
Brittanica or Empires of the Middle Ages, I don't think it's right that
someone who has a prosperous empire should feel obligated to throw it
all away attacking the leader because the victory conditions make no
distinction between finishing second and last--both are "losers"--or
because there's no absolute scale of accomplishment.
Although I haven't played it, I know that Third Reich is a game that
doesn't follow this pattern. From what I've read, it provides
independent victory conditions for each player, which makes possible
not only alliance wins in multi-player games, but even degrees of
victory on opposite sides. It's conceivable for Italy to minimize its
contribution to the Axis cause and end up with a victory even though
Germany goes down defeat and each of the Allies wins. The *possibility*
that Italy might do this should bring an interesting, complex, and
subtle dimension to relations between the Axis partners.
If we generalize this approach, we see that it makes possible a number
of diplomatic maneuvers that ordinarily make no sense in the typical
"zero-sum" game. As above, a partner that feels it is getting the short
end of the stick may plausibly threaten to cut a better deal with a
rival alliance. In most games as the end nears, the other partner will
simply laugh: "What, are they going to let you win to keep me from
winning?" Or, depending on the situation, the weaker partner may have
no choice but to switch sides, since the stronger partner has nothing
meaningful to offer. Alliances are just too brittle in a zero-sum
endgame.
Another maneuver is "divide and conquer": a powerful player can
plausibly break up a hostile grand alliance by bribing the weakest
members of the alliance, something that is really only possible if the
game either has independent victory conditions or if rank at the end of
the game has more than an infinitesimal value to the players. In this
case, the powerful player can say, "I will move you from 5th to (a
distant) 2nd place if you join me." Or, "Your allies can only give you
a Tactical Victory by giving up one of their Decisives, which they
aren't about to do. I will be able to give you a Tactical Victory if we
ally. Join me." (One should do one's best Darth Vader imitation when
uttering these words.)
Obviously this approach tends to decrease the overall tension in a
game. (Compare, e.g., a game of Diplomacy between seven people who all
think that an outright win is the only desirable outcome to one in
which some of the players players place a nonzero value on joint
victories, high rank, and/or survival. The latter will be much less
tense and alliances much more stable.) But it doesn't mean that a game
has to be a nicey-nicey demonstration that "if everyone would just
cooperate, we'd all be better off." The trick is simply to make the
overall game environment, in terms of supply centers, victory points,
economic resources, or what-have-you, sufficiently harsh that it can't
support a decisive victory for more than X out of the N players.
It may also be desirable in certain situations to create some exclusive
conditions, e.g.: "The Gormblats cannot do greater than a Tactical
Defeat if the Flugelheims achieve a Tactical or Decisive Victory." This
might represent a historical enmity so strong that those two players
have a truly zero-sum relationship. Similarly, one side might be
required to have a dominant position (highest number of victory points)
at the end of the game, in addition to fulfilling other, absolute
conditions, in order to claim victory. This would be appropriate for
the player of a large empire when all the other players represent
smaller countries or barbarian tribes. Imperium Romanum II uses a
variety of victory conditions, some of them absolute and/or
independent, some of them exclusive, to represent the political
situations in the various scenarios, something which I (obviously)
think is a strong point of that game.
--Elliot Wilen
But unless I'm mistaken (it's been a while), the Q&A in the
back of the rules explicitly clarifies that the use of the
Voice in combat MUST precede the use of Prescience. (It
is possible I'm remembering this in the wrong order -- This
was always one of our biggest points of "discussion" in the
Dune. :-)
And I definitely know you could not use Prescience to ask
if your opponent OWNS a card. You could only ask to see
one of your opponent's combat elements (attack, defense,
leader, or number dialed). If your opponent owns a shield,
but is not playing it, you would be none the wiser. Of
course that doesn't stop you from Voicing him to play a
shield anyway. If he doesn't have the card, he just ignores
you.
(Of course, a goomd Atreides player should definitely KNOW
whether or not his opponent holds a shield, unless it's the
Harkonnen or it's the original card dealt to a player. Or
you COULD always use Truthtrance to ask the "Do you have a
shield" question that you pose above. And I'm sure I remember
Truthtrances could be played ANY TIME during the combat process,
before, after, or during Voice or Prescience.)
We typically found that Voicing a player to PLAY a certain
card was not the optimal play, unless you knew for a fact
that that card was held (or if the BG was just fishing for
a Lazgun-Shield suicide run). About 99% of the time the BG
would Voice to NOT play the defense that matched her attack.
(In your example, if BG plays a Poison Weapon and is not certain
of opponent's cards, Voice should be "Play NO Poison Defense.
Let them waste the Shield if they have one and want to. This
"negative Voice" is also useful if you don't even have a
Weapon, but your opponent doesn't know that. Let them assume
you have their leader dead so he'll play a cheap one or something
like that.
This is not to say that the BG-Atreides alliance is not
one of the strongest. It certainly is. Most of the alliances
can be very, very powerful in the right conditions, but some
are always powerful. I believe that ANY alliance with the BG
is normally more powerful that any other, which I personally
think is the BG's main advantage. They can't do a whole heck
of a lot by themselves (unless they have an extremely wealthy
patron), but EVERYONE wants the Voice on their side. BG is
normally the first one sought out for an alliance in our
games, unless the front-runner needs cards and doesn't have
much money, in which case it's the Emperor.
What do you think are the most powerful Dune alliances
under normal circumstances?
A similar question: What do you think the most powerful
Dune powers are? Of course, all abilities have a place and
a time, but some are just inherently powerful plays.
My votes, in order:
1) played properly, the Harkonnen Hand Snatch
(optional-advanced Karama power)
2) the BG Voice
3) the Fill-the-Emperor's-Ally's-Hand-With-"Free"-Cards ploy
4) the Atreides Full-Battle-Plan-Prescience
(optional-advanced Karama power)
Later,
Brad Johnson (joh...@cig.mot.com)
Interesting. Just how do you track the Gnomes' money? Someone sitting ther
tabulating their income and expenditures? This isn't a flame or sarcasm, BTW,
but honest curiosity. We go with the rule that you don't have to show anyone
how much money you have, and one of the people who plays the Gnomes is *truly*
paranoid and cautious. I think he has won twice now with the Gnomes.
>Admittedly most of our games are fairly paranoid. A solo win by the
>Discordians when everyone had just expended all their effort stopping another
>player from winning, and a solo by the Bavarian Illumnati using their
>special ability are the only lone wins to date. The rest have been 2-way wins
>at best,
Interesting. We have had several wins, including wins from the UFO's (I agree
that they are a powerful group to have, if you know how to play deceptively
from the start) and the Bavarian Illuminati, even after their power requirement
was raised.
>John
>
Have to grab my friend who owns Illuminati once he gets back from his vacation.
:)
Cos
(John Costello)
cost...@ohsu.edu
We have a rule that you can't steal from the bank, but no such rule about giving
to the bank. Heh. That could be fun.
[Jeff Goldsmith stuff deleted.]
[Luke stuff deleted.]
> -Luke
>--
>______________________________________________________________________________
>Luke Hankins han...@cs.swarthmore.edu |"Le coeur a ses raisons que la raison
>for(MyAge=0;!Dead;EnjoyLife); | ne connait point" -Blaise Pascal
>While not (EndOfLife) do Havefun(me); |"...stay sane inside insanity..." -RHPS
>In article <CIrFG...@usenet.ucs.indiana.edu>
>aha...@bigbang.astro.indiana.edu (Alec Habig) writes:
>
> Most of the people I play with are competitive enough that they regard a
> shared victory as not really a victory at all, so this never happens.
> This sort of rule is my big gripe about the game Dune, where shared
> victories are about the only way to win.
>
>I strongly disagree about your assesment of Dune. I have played this
>game *many* times and have seen more solo victories than alliance
>victories, although admittedly alliance victories are still around
>1/4 to 1/3 of the outcomes.
I must admit, I've played at least 100 games of Dune over
15 years, and in that time I've only seen maybe <10 solo
victories, all by sheer fluke.
[...]
To extend the game and make solo victories more possible (and
attractive), we've tried house rules such as:
1) _Forbid_ alliances before turn X (usually X = 3-6).
2) _Require_ all alliances to be dissolved (forming
all new ones is allowed) at _every_ worm.
3) Require alliances to control 2+X cities to win
(where X = number of players in alliance).
I forgot that we were using your house rule #3. This makes a
*big* difference in play and makes alliance wins much less
frequent. In fact, I don't think we ever played the game
using the original alliance rules.....
The thoroughly scary Dune 2 players alliance is the Atreides - BG. How does
one stop an alliance that can virtually pick which person will get a leader
killed in battle each turn?
This is a powerful alliance indeed and in our games it usually provokes the
rest of the players to ally against it as a defensive measure and that
usually stops them dead in their tracks so this alliance is not often
suggested in our games. My main problem with Atr-BG alliance is that it
is really a drag for the other players to wait around while they figure
out their plan of attack ;-)
You can get sneaker than that. As BG ally with the person you didn't
predict to win. If the other side can be made to win in the turn you
predicted, you win. If not and you and your allies win, you win.
I pulled this one in a game, and unfortunately the game was won in
the following turn. At the end when I revealed my prediction, everyone
got very wary of me in following games.
Shawn Garbett
>je...@gg.caltech.edu (Jeff Goldsmith) writes:
>>The idea of making the Gnomes' income random is a good one, though.
>What you could do is take a deck of ordinary playing cards. Each turn,
>the Gnomes may take $12 or draw 2 cards from the shuffled deck. Count
>aces as '1' and face cards as '10'. These cards count towards the gnomes
>victory condition. The gnomes can cash in any of the cards at any time if
>he needs the money for something. Also allow collecting 1 card and $6.
>This will give the Gnomes an extra $.54 per card drawn on average.
If you take the nines and tens out of the deck beforehand, you get
each card worth exactly $6 on average.
--
Tim Shippert ship...@alumni.caltech.edu
Back off, man, I'm a _scientist_!
In article <CBARBER.93...@apricot-fddi.bbn.com>,
cba...@bbn.com (Christopher Barber) writes...
>One of my favorite solo vitories was actually through an alliance: I
>convinced everyone playing to ally against one other player who had
>been really annoying, so that he would be the only one to lose.
>However, I was playing the BGs and had predicted one of my allies to
>win in that turn, so I got the solo victory instead.
You can get sneakier than that. As BG ally with the person you didn't
predict to win. If the other side can be made to win in the turn you
predicted, you win. If not and you and your allies win, you win.
I pulled this one in a game, and unfortunately the game was won in
the following turn. At the end when I revealed my prediction, everyone
got very wary of me in following games.
Yup, I've done that. The BG is my favorite role to play in Dune mostly
because I enjoy trying to manipulate the game to fulfill my prediction.
I have won by prediction many times, although admittedly always within the
same group of guys (it's *much* easier when you know who you are playing
with). One really subtle thing you can do as the BG is to prevented someone
for playing for the win in a given turn by bluffing your prediction. This
is really hard to do since you need to somehow give the other guy the
impression that you predicted him without telling him directly (since then
he won't believe you).
BTW, how many out there have performed the BG "hat trick": predicting
yourself and fulfilling the prediction, thus winning twice over? I
have never tried it myself, but I have two friends that have each done
it once.
Need I mention that as BG allying with the person you DIDN'T
predict to win, if you can maneuver a big confrontration between
your alliance and the alliance containing the player you DID
predict to win on the right turn, one or two judicious thrown battles
can do wonders to get the solo victory. We had a BG player
who performed this masterfully once. (Once. :-)
Brad Johnson (joh...@cig.mot.com)
> I still have mixed feelings about Illuminati--I have a hard time playing a
> game that's both intense and serious in terms of alliances and backstabbing,
> but at the same time very fast, random, and general silly. If I want
> intense, I'll play something less random, like Diplomacy; if I want beer-
> and-pretzels, I'd rather go with something like Wiz War or Cosmic Encounter.
> Your mileage, however, may vary.
Anyone ever play the cheating rules for Illuminati? For those who
don't know, the idea of the cheating rules is that except for a few
special rules, anything is legal, but if you get caught doing something
that would normally be illegal, you have to undo it. Which rules are
*special* varies, but usually includes
(a) you can't physically touch or disrupt other people's power structures,
money stacks, or cards (otherwise the game falls apart, literally)
(b) you can't cheat when collecting your income (as this is extremely
difficult to verify)
(c) you can't smuggle in cards or money from another set (again, almost
impossible to stop)
I only played it once, it's a very strange game, particularly if you play
with a lot of people. You have to keep looking all around to make sure no
one else is getting away with anything, while trying to find opportunities
to get away with things. Ultimate beer'n'pretzel, but still entertaining
if you're into that sort of thing.
It turned out I lost because I forgot to keep track of people who weren't
playing (a spectator got a hundred out of the bank and gave it to someone)
but at least I got a pyhrric victory by having the Atlantean's seventh
Peaceful group in my pocket :) (the Antlanteans win if there are seven
Peaceful groups alive in the game, anywhere).
I usually find the biggest problem with Illuminati is trying to get
everyone to agree when someone is too close to winning and needs to
be stopped. There's always someone who doesn't think the danger is too
great and refuses to help, and so you either let the leader win, or stop
him and let the guy who didn't spend any effort to stop him win. Or,
one player "accidentally" puts most of his money off of his Illuminati
group, and so he simply can't put his fair share into the attack without
cheating. It's sort of a Prisoner's Dilemma type problem, where additionally
everyone is using a different rewards table.
> --
> Josh Smith, User Support Coordinator :: Pleasure: iri...@hmc.edu
I have the deluxe edition of Illuminati, and I've never seen the
atlanteans before. What other Illuminated Groups are there that are not
in the deluxe set, and where can you get them?
Thanks!
Pax,
--
------------------------------------------------------------------------------
Chris Taylor - ana...@netcom.com "Don't mock me, you're playing
Anarchy Nippon Software Company with fire!" - Tom Servo
Professional Game Designer/Semi-Professional net.junkie.insomniac
I may work for Interplay Productions, but my opinions work only for me.
Please send all Interplay related questions to: 76702...@compuserve.com
The Space Gamer (back when Steve Jackson Games published it) put out
several 'semi-official' additional Illuminati groups. Some favorites:
[Memory is a little fuzzy - my magazines are in a pile somewhere. ;-)]
The Illuminated IRS - Goal: Collect 5 Government Groups. Special
power: Ability to drain treasury of any one non-iluminati treasury
per turn (counts as an action).
The Illuminated Moral Minority - Goal: Collect 6 Conservative groups.
Special power: Immune to attacks from liberal and weird groups. (The
Anit-Discordians. ;-)
The Secret Society for the Abolition of Secret Societies: Goal - Collecting
groups with a total *resistance* of 20 (or something like that.) Special
Power: May disable any other Illuminati's special power for 5MB
The British (Pythonated) Illuminati - Goal: Collect 5 'Silly'
Groups[1]. Special Power: May not be destroyed - even if reduced to 0
groups after turn 3.
[1] This came with about 20 more groups based on Monty Python (e.g.
Ministry of Silly Walks, British Dental Association, The Spanish
Inquisition[2], etc.)
[2] The Spanish Inquisition had a *very* interesting special power of
its own; all attacks from the group are priviledged - Nobody expects
the Spanish Inquisition. ;-)
--
Capt. Gym Z. Quirk (Known to some as Taki Kogoma) kog...@unm.edu
I'll get a life when someone demonstrates it would be superior to
what I have now.
I don't know if I have the Atlanteans in my files, but I do have 3
files of new Illuminati groups available on my mail server.
Send a message to mail...@circle.raleigh.nc.us with the following in
the body of the message:
--- SNIP ---
help
index
--- SNIP ---
This will send you the help file for the mailserver and an index of the
files available.
Lance
--
Lance A. Brown <br...@cs.utk.edu>
The Crystal Wind is the Storm, and the Storm is Data,
and the Data is Life -- The Player's Litany
from _The Long Run_ by D.K. Moran
> I have the deluxe edition of Illuminati, and I've never seen the
> atlanteans before. What other Illuminated Groups are there that are not
> in the deluxe set, and where can you get them?
The deluxe edition didn't include anything from the third expansion
set to the plastic boxed edition (mostly since the expansion set was
sort of silly). It added some new rules, and some new Illuminati to
go with them. Here's a capsule summary (from memory, so forgive me
if I forget something).
Brainwashing: works like an attack to control (all modifiers and such)
except a successful attack changes an alignment of the target group by
one level (i.e., it takes two brainwashing attacks to take a group from
Straight to Weird). Illuminati groups get a free brainwashing attempt
at the beginning of the turn, and you can also take one as an action.
Helps out the alignment based Illuminati (Discordia, Bermuda, Assassins)
in particular. Obviously Illuminati groups themselves are immume.
Propaganda: the propaganda wheel included 9 tracks (for each alignment
but Fanatic) with opposite alignments opposite each other. Five markers
are used to indicate the level of world opinion, and you can move these
back and forth to make the world more weird, less violent, whatever. As
alignments become more favored, groups which have those alignments get
a bonus to their power and income (up to a maximum of +3). Also, if the
marker gets to the end of the track (6 spaces) then not only do groups
with that alignment get the +3 bonus, groups of opposite alignment get
a -1 to both. You get a free propaganda attack at the end of your turn,
or can take one as an action.
New Illuminati:
Atlanteans: moderate power and income, as I recall. Their special
victory condition is to have 7 peaceful groups *in play*, anywhere
(but not counting dead groups). Also, each violent group they destroy
counts one. Since there are only 7 peaceful groups in the deck, obviously
this doesn't work very well without Brainwashing. I think their special
power was a bonus to brainwashing.
Voodoo somethings: pretty good power and income, special power is that
all propaganda attacks are priviledged. Special victory conditions is
to choose 2 alignments at the beginning of the game (like the UFO's)
and get world opinion to the top of those two alignments on the propaganda
chart. Plays much like the UFO's, obviously requires the propaganda rules.
There was another Illuminati, the only thing I can remember about it
was that it's special victory condition was to get groups with a combined
resistance of 30.
The main effect was to make the game a lot longer without adding much
(longer since you have more to do each turn). Once the novelty wears
off, it's easier to just play the basic game with the first two expansion
sets (all of which are incorporated into the deluxe edition).
To get these rules, you have to find the third expansion set, and I have
no idea if it's still in print. The SJG people would know.
> Thanks!
> Pax,
> --
>Interesting. Just how do you track the Gnomes' money? Someone sitting ther
>tabulating their income and expenditures? This isn't a flame or sarcasm, BTW,
>but honest curiosity. We go with the rule that you don't have to show anyone
>how much money you have, and one of the people who plays the Gnomes is *truly*
>paranoid and cautious. I think he has won twice now with the Gnomes.
Precisely that, yes. One of teh players sat down and kept careful count. Of
course, he was much more careful about keeping track of income than of
expenditure, so that by the end of the game the Gnomes only had 40MB whereas
the balance showed 80MB. :). He made a convenient source of cards, as people
would tend to chip in on attempts to control, just to weaken the Gnomes.
>>Admittedly most of our games are fairly paranoid. A solo win by the
>>Discordians when everyone had just expended all their effort stopping another
>>player from winning, and a solo by the Bavarian Illumnati using their
>>special ability are the only lone wins to date. The rest have been 2-way wins
>>at best,
>Interesting. We have had several wins, including wins from the UFO's (I agree
>that they are a powerful group to have, if you know how to play deceptively
>from the start) and the Bavarian Illuminati, even after their power requirement
>was raised.
Admittedly the main reason for lots of joint victories is time considerations,
or to avoid losing to another player as a result of stopping player one from
winning.
John
I've played it, and didn't like it that much. Maybe our group was just far to
energetic at cheating, but it gets really tiresome to have to mentally
follow along with absolutely every detail in every calculation made in the
game. Maybe if you added the folowing rule -
(d) anyone caught cheating must pay a penalty of 5MB or release control of a
group.
That way people would only try to cheat when they thought they could get away
with it and it was really worth it, instead of constantly.
Alec
Cool Cheats:
Tossing in several money chits into the bank to affect the odds ... and
misstating how much it was
Having someone *else* slide you extra MB from the bank for income and/or
change.
Keeping a group drawn as if it were a special card, then using it face-down
for a priviledged attack.
Misadding power/resistance aloud : "Okay, I have a power of 9, plus 2 transferred from [x], plus similar alignments for 4 and 4, and I'll throw in 6, against
a resistance of 7 is 20, less 10 for proximity to 'Nati group is a ten or less."
Rolling dice quickly, without pause, so people can't throw in MB against you.
Just a few I've used (or been victim of!)
... Hurrican George _can_ work !! ...
-CS
PS: The best cheat, however, is <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> <fnord> . But you already knew that, right?
On Compuserve some PBM rules have been posted that some people are using to
play quite effectively. The Honor system is still used to regulate play,
which is the primary drawback I see with playing for an ante (no one wants
to lose their favorite card to someone who may or may not be honest). This
problem may have no simple resolution. I would suggest maybe a clearing
house for PBM players. The rules posted by ARGUS on Compuserve involve
compiling a numbered list of the cards you are going to use. Your opponent
then tells you what number card will be your draw. Since your opponent
doesn't know how you have numbered your cards, the draw is essentially
random. When the game is over you then, theoretically, exchange lists to
verify play. This still does not stop a player from fabricating a numbered
list after the game is over. What would work, however, is if each player
developed a numbered list and then faxed, posted, or e-mailed that list to
a neutral third party observer for safekeeping: this is the Clearinghouse
concept. The Clearinghouse then reposts the lists to each opponent after
the game has been concluded. This would work as long as the Clearinghouse
is compensated in some way for their trouble. That is a whole issue
altogether. I suppose supply and demand will determine the cost of this
service. Hmmm.... maybe I'll work on that.
Fred
Now, you just gotta trust the other guy not to crack your password :)
--
Eu-Ming Lee mi...@interaccess.com
"i showed this girl my stitches
"she said she had some too
"she said she thinks she'll start a rock band too." - Throwing Muses
Well, just to throw in a counter opinion, I really think the third
expansion added a great deal to the game. I just don't enjoy Illuminati
without the third expansion anymore, as I find the game boring and painfully
predictable. The reason for this is that I really feel the third expansion
set gave you alot more options that ultimately gave you alot more control
over the game.
Once you get used to it, I really don't think games with the third expansion
set take all that much longer than games without.
>To get these rules, you have to find the third expansion set, and I have
>no idea if it's still in print. The SJG people would know.
It is very out of print and, I virtually never see them for sale from used
game dealers or in auctions. So, if you do see one you might have to pay
quite a bit to get it. (Now watch as a half dozen net.people offer theirs
up for sale.)
Nick Sauer
>Are there any rules I overlooked that can slow the Gnomes down? I read
>through the book today, but didn't find anything. It just seems as if
>the Gnomes can make a normally dynamic game a real bore as the Gnome player
>takes a very stoic, slow burn approach in a normally fast moving, dynamic
>game. We decided to ban the Gnomes in the future unless we could decide
>upon an alternate winning condition.
>Any ideas on what to do with them?
>Thanks,
>Matt
>mx...@po.cwru.edu
well, i never played the older version, but there is a rule (at least in the
new version) that if any time after the third turn (i think, been a while
since i played) if a group has no groups, it is out of the game. so if
the gnomes are being as lame as you indicated, just kill off all his groups
>:)
just the threat of such an attack oughtta jode him into being more active
-the count
>Anyone ever play the cheating rules for Illuminati?
Yes - in fact, back in the old days when the pocket box edition
first came out, we ONLY played the cheating rules. The first time I played
a "straight" game was when I went to college and couldn't convince others
to try it the REAL way.
>(a) you can't physically touch or disrupt other people's power structures,
>money stacks, or cards (otherwise the game falls apart, literally)
I believe we played you could steal the cards in their hand, but
the power structures and money were verboten.
>(b) you can't cheat when collecting your income (as this is extremely
>difficult to verify)
>(c) you can't smuggle in cards or money from another set (again, almost
>impossible to stop)
Hey, there IS honor amongst thieves! We never had any problem with
people violating the sacrosanct rules. What's the point in playing if you
are going to REALLY cheat?
You forgot the most important rule, though - if you actually got
caught cheating, you had to put things back the way they were (but with no
other penalty). We interpreted this to mean that you had to be able to
tell exactly when the cheating thing happened in order to make the cheater
fix it. So if you found a new group in one person's structure that you
knew he hadn't gotten fairly, you couldn't do anything about it unless you
could say when he put it there (which generally meant you had to catch him
in the act). This lead to a lot of ritual accusations, like calling anyone
when he kept a card or when they calculated the attack number, etc, but by
and large it worked fairly well.
>I only played it once, it's a very strange game, particularly if you play
>with a lot of people. You have to keep looking all around to make sure no
>one else is getting away with anything, while trying to find opportunities
>to get away with things.
There are different styles of cheating, and different players will
rapidly get different reputations and be subjected to differing levels of
surveillance (because you can't watch *everybody* *all the time*). One of
my friends was a serious kleptomaniac - we took to slapping his hands when
he tried to steal a card from the deck. By the end of the game his hands
would be red, but as he would say, for every five time we caught him he'd
get away with it once, and we caught him a _lot_...(BTW, this friend's
reaction when I told him about the bridge sized cards in the deluxe version
was, "Those would be a lot harder to palm.")
I, on the other hand, had a reputation (somewhat undeserved) for
not cheating very often, and therefore the others wouldn't watch me too
closely. I would have a much better chance of getting away with something
when I _did_ try it, a situation I exploited several times to win games by
less than pure means.
I would agree with the suggestion in the rules that you should play
this game with friends rather than mere acquaintances, and nobody should
take it too seriously. Even so, it can get really intense as you try to
keep your eyes on all five of your opponants at once while trying to figure
out a way to sneak that card you palmed into your structure. I don't
believe I've ever played another game that required so much continuous
attention; its really quite an exhausting (yet exhilarating) experience.
--
Tim Shippert ship...@alumni.caltech.edu
"Not all Democrats are smart, but all Republicans aren't."
-John Kenneth Galbraith
>Atlanteans: moderate power and income, as I recall. Their special
>victory condition is to have 7 peaceful groups *in play*, anywhere
>(but not counting dead groups). Also, each violent group they destroy
>counts one. Since there are only 7 peaceful groups in the deck, obviously
>this doesn't work very well without Brainwashing. I think their special
>power was a bonus to brainwashing.
In my german box set of Illu the win condition is 5 groups. Normally we
remove the Atlanteans because our group does not like random winners :-)
Heiko
And the Atlantean special power is to double the income of any peaceful
group they control, very nice if you can brainwash, but not too good otherwise.
- Paul
> The group I played Illuminati with at Swarthmore actually had the opposite
> problem with the Gnomes: no one ever wanted to play them, because everyone
> would always gang up on them. The strategies others have described make it
> virtually impossible for the Gnomes to succeed, especially when coupled with
> a particularly nasty tactic someone came up with before the last game in
> which someone voluntarily played the Gnomes.
>
> This player, let's call him Bruce (hi Bruce :^), kept detailed notes of
> exactly how much money the Gnomes had at any given time. Every time they
> earned income or spent any money, Bruce jotted it down on a piece of paper,
> so everyone knew exactly how well (or poorly) they were doing... And
> whenever Zurich got too high in the minds of the other players, they'd just
> take turns whacking off groups until the Gnomes were forced spend money to
> defend themselves (or be eliminated, by the at-least-one-group rule).
>
> This was very effective at stopping the Gnomes, but very frustrating for the
> Gnomes' player, so much so no one in that group ever chose the Gnomes again.
If I were playing the Gnomes, and someone tried to do this to me, I would
simply inform that player that if he did so, he would never win the game.
Period. I don't believe any Illuminati can win if another is completely,
constantly, utterly opposed to whatever the first one does. See just how
badly you can beat him up! And if he even threatens to win, well, you can
always give the right groups to a third player....
I'd do much the same thing if all the other players ganged up on the Gnomes.
A single player can't be picked out for retaliation, but if you're careful to
acquire groups the others could use and not let your treasury get so large
as to attract more abuse, you're set to make someone else into a juggernaut.
Preferably someone in third or fourth place, of course! After a couple of
games in which playing well just gets you blindsided by the Gnome-giveaway,
I think the anti-Gnome fervor would die down.
This is, of course, metagaming, and a lot of people don't like that. But
I think it's wholly justified, for two reasons. First, I'd like to point
out that keeping track of the Gnomes' finances is itself metagaming,
particularly when you use pencil and paper. The game doesn't come with
those, so why do you think you can use them? Now if you can keep track in
your head, more power to you--but while you're tracking my finances, I'll
be figuring out how to demolish your power structure. And if you won't
metagame in the first place by keeping track, I won't have to metagame
in retaliation.
The second reason is that the point of playing the game is to have fun; by
using this tactic, one or more players are deliberately preventing the
Gnomes from having fun, at least in terms of winning the game. Well, what
you can't get one way you can usually get another, and while throwing the
game to someone else isn't as much fun as winning yourself, it's certainly
entertaining! Particularly when the other players don't think it's fair....
Dave Cole
d...@otto.uchicago.edu
[lots of stuff deleted]
>The second reason is that the point of playing the game is to have fun; by
>using this tactic, one or more players are deliberately preventing the
>Gnomes from having fun, at least in terms of winning the game. Well, what
>you can't get one way you can usually get another, and while throwing the
>game to someone else isn't as much fun as winning yourself, it's certainly
>entertaining! Particularly when the other players don't think it's fair....
Presumably by the same measure, you don't keep track of how much power
the Bavarian Illuminati have, or how many wierd groups...
The Gnomes victory condition is a pile of megabucks. Its just good playing to
have some idea of how close the various powers are to winning.
>Dave Cole
>d...@otto.uchicago.edu
John
: [lots of stuff deleted]
: >Dave Cole
: >d...@otto.uchicago.edu
: John
I think there is a difference between keeping track of the Gnomes' (or anybody
elses) money and checking power, weird groups etc. The power score is easily
checked *without* writing down or memorizing numbers or other information. As
I remember you can't hide your groups. Keeping track of money means writing
down the numbers (at least for most people). I wouldn't want anyone to do this
evan if I wasn't the Gnomes. I wouldn't do it myself unless, maybe, against
someone who had done it himself...
-Anders
-Anders
I dunno, World in Flames (at least 5th edition) has optional individual
victory conditions. Personally, although I haven't played Third Reich, I
don't think I'd like it as well due to the relative lack of detail in the air
and naval systems. I also like WiF's treatment of the entire war, not just
the European Theatre. Just depends on what you think is important, I guess.
--
-Doug Gibson d...@wiffin.chem.ucla.edu
"I didn't like the way he was bleeding, so I made him stop."
-from Mutant League Football
>The second reason is that the point of playing the game is to have fun; by
>using this tactic, one or more players are deliberately preventing the
>Gnomes from having fun, at least in terms of winning the game. Well, what
>you can't get one way you can usually get another, and while throwing the
>game to someone else isn't as much fun as winning yourself, it's certainly
>entertaining! Particularly when the other players don't think it's fair....
I think this transcends "metagaming" and reaches into
"poor sportsmanship." We have a firm rule for all games
that it is not permissible to try to choose someone else
to win rather than yourself. It's too easy to help someone
else win.
--Jeff
--
You think I'm bidding with that Yarborough there?
Hearts and clubs and diamonds and spades...
And not one card that's over a nine,
You once were a partner of mine. --Colin Green
>d...@otto.uchicago.edu (Dave Cole) writes:
>[lots of stuff deleted]
There is a difference. Power is public knowledge. Money is expressly
claimed to be secret by the rules. Keeping exact track of someone
else's money contravenes the rule of secrecy. In any case, it is
clearly illegal to keep track of someone else's money *and then
tell another player* your findings. That information is secret;
it is not permissible, therefore, to transmit it.
There was a somewhat parallel discussion on GEnie recently, regarding
the game Breakout: Normandy. Combat is normally resolved by each player
rolling two dice; however an alternative method is for each player to
use a set of 36 chits (one chit labelled "2", two chits labelled "3",
and so on) and draw one at a time until exhausted. This obviously makes
the rolls "average out" over the course of the game; however, it means
that an attentive player can count out the chits and always know what
the 36th combat result will be.
Some players dislike chits for this reason; some players just accept
counting as part of the game. It turns out that the designer, Don
Greenwood, likes to play with chits, but strongly dislikes counting.
His opinion is that the players (even when playing by mail!) should play
with chits but simply agree not to keep track of the draws. Or, in the
alternative, if you don't trust your opponent not to count, you should
revert to dice instead.
Another related discussion involves Advanced Civilization, where the
official rules for reshuffling the deck after each round make the
positions of the untradeable calamities known to everyone, and thus easy
to count in order to try to avoid. Again, I thought that this was an
obvious and natural consequence of the rules, but it turns out that
Bruce Harper, the designer, considers counting unsportsmanlike.
Yet another discussion comes up in tournament play of Acquire, in which
both cash and stock holdings are often considered "secret," although
they can of course be counted. In a recent General a successful Acquire
player actually advocated modifying ones strategy to buy stocks in "odd
batches" simply in order to make counting harder.
My personal opinion is that I play games as a test of skill, and that
there is no point in testing my ability to count, since anyone who wants
to do can do that. So I consider games which have "secrets" which are
easily broken by counting to be rather poorly designed. I know a lot of
other players who agree with this, and make house rules that anything
that can be kept track of is public. This avoids both the pointlessness
of turning what should be an interesting activity into a counting
exercise, and the futility of trying to make rules to regulate the
players' thought processes.
> In any case, it is
> clearly illegal to keep track of someone else's money *and then
> tell another player* your findings. That information is secret;
> it is not permissible, therefore, to transmit it.
I think that you are in a small minority of players in drawing this
conclusion. There is an analogous situation in Titan: if you are
attacked and flee, you learn the contents of an enemy legion, but the
rules state that this information is secret from the other players. Of
approximately 10 people I asked about this subject, only one thought
that this should mean that the player with the information is forbidden
to reveal it to other players; all of the others thought that a player
can say whatever one wishes, and that the stated rule only implies that
a player cannot _show_ the enemy legion to other players.
David desJardins
--
Dean R Biron, University of New Hampshire Research Computing Center -><-
Prefector of the order of St Gulick, Legion of Dynamic Discord -><-
I'm so tired of all the f-cked up minds of all the terrorist religions and
No, not quite. The rules explicitly mention that the money in a group's
treasury is a secret, except for what is visable on top. The power structure
of a player is certainly not secret. In my mind, this is part of the Gnome's
victory condition - no one knows exactly how close he is - just like no one
knows the UFO's VC (which is also expressly mentioned as secret). A good Gnome
player should be artful with making change and distributing his money - after
all, his special power is to freeley rearrainge his money. Should make it
harder to trace, unless someone is going to be really anal about it and play
Mr. Tax Accountant. If I were the Gnomes, and someone tried that on me, I'd go
out of my way to mess him up too (repaying his actions towards me).
Alec
> No, not quite. The rules explicitly mention that the money in a group's
> treasury is a secret, except for what is visable on top.
Certainly the money in any single group's treasury is a secret, since you
aren't allowed to search through each treasury. I've always assumed that the
Gnome's special power is equivalent to gathering all their money in one
pile, then making change without revealing what denominations are involved
(although the player should be very careful to not make mistakes here) and
then placing pieces of the pile whereever they want in their structure
without declaring what's going where.
However, this is different from knowing how much money is in *the whole
structure*.
> The power structure of a player is certainly not secret. In my mind, this
> is part of the Gnome's victory condition - no one knows exactly how close
> he is - just like no one knows the UFO's VC (which is also expressly
> mentioned as secret).
However, the total income and expenditures of the Gnome player are not
secret, which means that the difference is hardly secret either. I have a
real problem with saying that people aren't allowed to know and use the past
history of the game, particularly when this isn't an artifact of the game
structure (like the example David gave of drawing chits to determine a
random result) but is simply the moves that each player is making. Unless
you argue that knowing how much people are making and how much they are
spending isn't essential knowledge to the game, then I can't accept that
a player shouldn't keep track.
To look at it another way, if you accept that players should have some idea
of how close the Gnomes are, where do you draw the line between acceptable knowledge and knowing too much, how much effort are players allowed to put
into it? Is it fair to keep track of the number of turns the Gnomes net
+10MB or more?
In my mind, when information is available, the only choices are to
completely ignore the information (like ignoring which chits have been
drawn so that you maintain the illusion of random results) or allow full
use of the information. I can't accept the idea of allowing fuzzy awareness
while disallowing perfect knowledge, because I wouldn't know when I'm
supposed to stop calculating.
I don't see that it's practical to completely ignore how much money the
Gnomes have (you're supposed to have some idea) and so I have to swing
the other way, that the total cash flow is public information and you
can do whatever you like with that information. It's nothing more than
knowing what has happened, and reminding people of it.
> Alec
: >I recently bought Third Reich and are in the middle of a solo test run. I find
: >it very enjoyable, partly because of the victory rules mentioned in the
: >preceeding post. The single feature that I like most is that when there are
: >three or more players, Russia has her own, separate victory conditions. This is
: >a great improvement over World in Flames, my former favorite WWII game. Even
: >though Russia (almost) certainly will fight the Axis, she may not do so when it
: >suits the other Allies best.
: I dunno, World in Flames (at least 5th edition) has optional individual
: victory conditions. Personally, although I haven't played Third Reich, I
: don't think I'd like it as well due to the relative lack of detail in the air
: and naval systems. I also like WiF's treatment of the entire war, not just
: the European Theatre. Just depends on what you think is important, I guess.
What I like most about WiF is Days of Decision, when you can affect the pre-war
policies. Of course, playing the whole war is better than just the European
part. I have the 4th edition of WiF, and I don't remeber the individual
victory conditions, but it's possible that I've missed them.
If you want a detailed game then WiF is definitely better than 3R. I will
most probably play WiF in the future, but 3R fills the need for a game that is
somewhat quicker and easier to learn, but not as simple as, say, A&A.
-Anders
> I don't see that it's practical to completely ignore how much money the
> Gnomes have (you're supposed to have some idea) and so I have to swing
> the other way, that the total cash flow is public information and you can
> do whatever you like with that information. It's nothing more than
> knowing what has happened, and reminding people of it.
Also, drawing on a real world analogy. Think about how far companies go to
maintain the secrecy of their financial info and how far they will go to
get that info about their competitors.
Allowing people to track the income/expenditures of the Gnomes only forces
the Gnomes to act. I think this is good.
Lance
Agreed with you analysis of the legality of keeping track of the gnome's net
income. My point is that I, as the gnome player, realize that some degree of
monetary confusion is to my advantage. If a player wants to try and take that
away by being anal and informing the world (as opposed to just having a general
idea of what's going on), I reserve the right to ogg his socks off to make him
stop.
Not that the gnomes should ever expect to win a 150MB victory. If they've
gotten to only 2/3 of that level, they can pretty much buy the standard victory
anyway.
Alec
>>The Gnomes victory condition is a pile of megabucks. Its just good playing to
>>have some idea of how close the various powers are to winning.
>There is a difference. Power is public knowledge. Money is expressly
>claimed to be secret by the rules. Keeping exact track of someone
>else's money contravenes the rule of secrecy. In any case, it is
>clearly illegal to keep track of someone else's money *and then
>tell another player* your findings. That information is secret;
>it is not permissible, therefore, to transmit it.
The way my friends and I play games in general, info that's "secret"
just means that if you ask me, say, how much $ I have in Illum, I don't
have to tell you; we've got a house rule that if you're not actually
playing you can't give any info about another player's position, whether
it's public knowledge or "secret", but if I'm playing there's nothing to
stop me from telling you how much $ another player has. OTOH, if I want
to know something that's public knowledge, you can't hide the info.
-DarkMage
-- Gary
Subject: Re: Bone alliance wins (was: Illuminati: Gnomes of Zurich?)
From: cthu...@aol.com (Cthulhoid)
Date: 11 Mar 1994 23:45:01 -0500
Message-ID: <2lrhcd$m...@search01.news.aol.com>
Wow, other Bone enthusiasts exist in the world! I have just joined this
newsgroup, and I cannot tell if my messages are getting posted, since they
don't show up at my end. If some kind Bone player out there would send a
message to cthu...@aol.com confirming that you guys can read this, I would
be much obliged!
KULL WA-BONING!
-- Gary
>I vastly prefer games with sophisticated supply rules, as well as games with
>multiple-step combat units. Games lacking these features, unless they are at
>the grand-strategic level, simply cannot represent the attrition and supply
>concerns of real commanders in 20th-century campaigns.
Maybe so, but there is something to be said for games that can be played in an
evening.